AI Paper Atlas

Research map / Image generation / GANs

Deep Neural Network Augmentation: Generating Faces for Affect Analysis

Dimitrios Kollias, Shiyang Cheng, Evangelos Ververas, Irene Kotsia, Stefanos P. Zafeiriou

IJCV 2020 · Image generation › GANs · approach: GAN-based image manipulation · 114 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train on faces synthesized by the proposed 3D-model pipelineVGG-FACE-GRU · Aff-Wild valenceCCC: 0.570 -> 0.595Improved
train on faces synthesized by the proposed pipelineVGG-FACE-GRU · Aff-Wild arousalMSE: 0.060 -> 0.051Improved
augment training with proposed synthesized imagesResNet-GRU · RECOLA valenceCCC: 0.462 -> 0.554Improved
augment training with proposed synthesized imagesResNet-GRU · RECOLA arousalCCC: 0.209 -> 0.312Improved
augment training with StarGAN-synthesized images instead of proposedResNet-GRU · RECOLA valenceCCC: 0.554 -> 0.503Hurt
augment training with GANimation images instead of proposedVGG-FACE · AffectNet valenceCCC: 0.62 -> 0.56Hurt
augment training with 2,476,235 proposed synthesized imagesVGG-FACE · AffectNet arousalCCC: 0.37 -> 0.54Improved
augment training with proposed synthesized imagesVGG-FACE-GRU · AFEW-VA arousalMSE: 0.247 -> 0.075Improved
augment training with proposed synthesized imagesVGG-FACE · RAF-DB basic expressionsaverage diagonal accuracy: 0.589 -> 0.775Improved
augment training with proposed synthesized imagesVGG-FACE · AffectNet basic expressionstotal accuracy: 0.52 -> 0.60Improved
augment training with proposed synthesized imagesVGG-FACE · AffectNet basic expressionsF1 score: 0.51 -> 0.59Improved
augment with StarGAN images instead of AffWildNet landmark inputVGG-FACE-GRU · Aff-Wild valenceCCC: 0.570 -> 0.556Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy